Dear all,
I need to use reactive force field to simulate my collagen solvated in water. After learning lammps for one week, I saw this sentence “ The original TIP3, TIP4P, and SPC model [1] must be used with the pair_style lj/cut/coul/cut or pair_style lj/coul/cut command.” In this website: https://sites.google.com/a/ncsu.edu/cjobrien/tutorials-and-guides/working-with-water-in-lammps
Does this mean that reactive force field in LAMMPS cannot simulate system including water?
Any suggestions is appreciated.
Regards,
Ming
ReaxFF can’t simulate water with a TIP3P ff. It can
model it with a ReaxFF force field.
Steve
ReaxFF can’t simulate water with a TIP3P ff. It can
model it with a ReaxFF force field.
It is also advisable to check the literature for how ReaxFF works in general and whether parameters suitable for your system exists. The transition from conventional classical force fields to something like ReaxFF is non-trivial and comes with a significant increase in computational cost.
You will need much more than a week to have a good enough understanding of the implications of using ReaxFF. You are strongly advised to start with test calculations of very small systems.
Axel
Dear Axel and Steve,
Thanks for your suggestions and guidance.
I need to use reactive force field to simulate my collagen solvated in water, as I want to investigate the large deformation mechanisms. I want to fix one end of the collagen triple helix (36 nm in length containing 126 amino acids), and use fix smd to pull the other. I want to figure out whether my simulation is feasible in LAMMPS first. So I read every of the fix smd posts in the archive, but did not find any of them using reactive force field. Can you please help to tell me whether reax ff in LAMMPS is feasible to simulate my protein solvated in water system (around 3000 protein atoms)? Is there any easy way to build water molecule in data file ? I am new in LAMMPS. I having been using GROMACS. But it does not have reactive force field, so I change to LAMMPS.
Thanks for your help
Ming
Dear Axel and Steve,
Thanks for your suggestions and guidance.
I need to use reactive force field to simulate my collagen solvated in water, as I want to investigate the large deformation mechanisms. I want to fix one end of the collagen triple helix (36 nm in length containing 126 amino acids), and use fix smd to pull the other. I want to figure out whether my simulation is feasible in LAMMPS first. So I read every of the fix smd posts in the archive, but did not find any of them using reactive force field. Can you please help to tell me whether reax ff in LAMMPS is feasible to simulate my protein solvated in water system (around 3000 protein atoms)? Is there any easy way to build water molecule in data file ? I am new in LAMMPS. I having been using GROMACS. But it does not have reactive force field, so I change to LAMMPS.
You first need to figure out whether ReaxFF has parameters
Dear Axel and Steve,
Thanks for your suggestions and guidance.
I need to use reactive force field to simulate my collagen solvated in water, as I want to investigate the large deformation mechanisms. I want to fix one end of the collagen triple helix (36 nm in length containing 126 amino acids), and use fix smd to pull the other. I want to figure out whether my simulation is feasible in LAMMPS first. So I read every of the fix smd posts in the archive, but did not find any of them using reactive force field. Can you please help to tell me whether reax ff in LAMMPS is feasible to simulate my protein solvated in water system (around 3000 protein atoms)? Is there any easy way to build water molecule in data file ? I am new in LAMMPS. I having been using GROMACS. But it does not have reactive force field, so I change to LAMMPS.
My response is the same as before. You first have to find out, whether there are suitable parameter sets for your system and whether some similar study has been done with ReaxFF before. This has nothing to with lammps itself. And requires a thorough review of the relevant literature.
And carefully follow the rest of my suggestions. Using ReaxFF correctly requires more skills than running gromacs.
Axel
Thanks Axel,
I will do literature review, and will give you feedback then.
Dear Axel,
In Buchler’s paper (http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=8029117&fileId=S0884291400083710 ), he says “We use a particular type of the ReaxFF potentials as suggested in Refs. 38 and 44 with slight modifications to include additional QM data suitable for protein modeling.48” But I found that Ref.48 is not published. I want to use exactly the same method with Buchler’s to study callagen. I thought maybe this function has been put into reax FF, so I contacted its developer. It turns out that SMD is not applicable in ADF/ReaFF. But I found buchler’s another paper (http://www.sciencedirect.com/science/article/pii/S175161611400201X ), in which he says “The mechanical behavior of the cross-links (consisting of amino acids) has been assessed by steered molecular dynamics in LAMMPS (Plimpton, 1995). Using ReaxFF reactive force field, we are able to evaluate the mechanical behavior of the cross-links until failure (Fig. 2).” But it does not tell whether the protein is solvated in water during the SMD simulation. I contacted the author, but have not get any response yet. Do you have any suggestion for me ?
Also, I saw different ffield.reax for objects studied in example folds. Does this mean that I need to write suitable ff.reax (suitable parameters as you mentioned) for my collagen?
Thank you,
Ming
Dear Axel,
In Buchler’s paper
(http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=8029117&fileId=S0884291400083710
), he says “We use a particular type of the ReaxFF potentials as suggested
in Refs. 38 and 44 with slight modifications to include additional QM data
suitable for protein modeling.48” But I found that Ref.48 is not published.
I want to use exactly the same method with Buchler’s to study callagen. I
thought maybe this function has been put into reax FF, so I contacted its
developer. It turns out that SMD is not applicable in ADF/ReaFF. But I found
buchler’s another paper
(http://www.sciencedirect.com/science/article/pii/S175161611400201X ), in
which he says “The mechanical behavior of the cross-links (consisting of
amino acids) has been assessed by steered molecular dynamics in LAMMPS
(Plimpton, 1995). Using ReaxFF reactive force field, we are able to evaluate
the mechanical behavior of the cross-links until failure (Fig. 2).” But it
does not tell whether the protein is solvated in water during the SMD
simulation. I contacted the author, but have not get any response yet. Do
you have any suggestion for me ?
please remember that i am not your adviser. everything in this
paragraph is a discussion that you need to have with your adviser, not
any of us.
Also, I saw different ffield.reax for objects studied in example folds. Does
this mean that I need to write suitable ff.reax (suitable parameters as you
mentioned) for my collagen?
kindof. as i had already mentioned, you need parameters specific for
your system. for more information, you should read the documentation
on the reax/c pair style.
axel.
Dear Ming,
generating suitable parameter files for ReaxFF is not an easy task. To get meaningful results the parameter set you use hast to be fitted to the problem you are going to study. As mentioned in your posting it usually means you have to fit the parameters to QM calculations of the reactions you want to study.
In Buchler’s paper (http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=8029117&fileId=S0884291400083710 ), he says “We use a particular type of the ReaxFF potentials as suggested in Refs. 38 and 44 with slight modifications to include additional QM data suitable for protein modeling.48” But I found that Ref.48 is not published. I want to use exactly the same method with Buchler’s to study callagen. I thought maybe this function has been put into reax FF, so I contacted its developer. It turns out that SMD is not applicable in ADF/ReaFF. But I found buchler’s another paper (http://www.sciencedirect.com/science/article/pii/S175161611400201X ), in which he says “The mechanical behavior of the cross-links (consisting of amino acids) has been assessed by steered molecular dynamics in LAMMPS (Plimpton, 1995). Using ReaxFF reactive force field, we are able to evaluate the mechanical behavior of the cross-links until failure (Fig. 2).” But it does not tell whether the protein is solvated in water during the SMD simulation. I contacted the author, but have not get any response yet. Do you have any suggestion for me
Also, I saw different ffield.reax for objects studied in example folds. Does this mean that I need to write suitable ff.reax (suitable parameters as you mentioned) for my collagen?
At Matcalc we have experience in fitting parameter sets to special problems. If you have any questions about generation of such a parameter set you can contact me.
Stephan
Reax has its own forcefield that includes H and O and it can be used for simulations with water. You should not use any other water potential other than the H and O potentials that accompany reax.
I’d suggest you read the papers by Adri van Duin et al prior ro performing your simulations to get a better knowledge of Reax.
Jim
James Kress Ph.D., President
The KressWorksTM Foundation ©
An IRS Approved 501 ©(3) Charitable, Nonprofit Organization
“KressWorks: Engineering The Cure” TM
(248) 605-8770
Learn More and Donate At:
http://www.kressworks.org
Confidentiality Notice | This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential or proprietary information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, immediately contact the sender by reply e-mail and destroy all copies of the original message.
Dear Stephan,
Thanks for your kind offer.
Recently, I got another problem. My system consists of protein solvated in water. How can I generate the input data file used for LAMMPS simulation?
Best regards,
Ming
See doc/Section_tools.html and this link:
http://lammps.sandia.gov/prepost.html
Steve
Hi Steve,
Thank you. This link is really helpful.
Cheers,
Ming